I am starting to experiment with AI on this site. This is the process I use to write articles using AI.
- Voice-to-Text Capture
I use my cell phone as my primary tool for capturing thoughts—because it’s portable and I can take it anywhere, like to Starbucks. I rely on Gboard’s voice-to-text feature, which works well with an external microphone to improve accuracy. I speak freely—sometimes off the top of my head, sometimes referencing notes—without worrying about grammar, structure, or repetition. The external microphone let’s me hold the phone so I can watch the text as it appear in real time so I can see if I need to speak more clearly or louder or slower. And it gives me the option to correct any significant mistakes immediately. This lets me get my raw ideas down quickly and efficiently. The more thoroughly I can explain my thoughts in this step, the better the resulting outline and first draft are. - Initial Cleanup
After recording, I go through the transcribed text and clean up obvious errors—like words that were misheard, incorrect homophones, or spoken corrections. I also fix any awkward phrasing or redundant thoughts. I don’t worry about organizing the content or removing repetition at this stage—my focus is only on clarity and accuracy. I want the AI to work with a clean version of my ideas. - AI Processing
Once the text is cleaned up, I copy it into an AI tool—right now, I’m using Leo on the Brave Browser. I chose it because it’s free, reliable, and, most importantly, it has a strong privacy policy: it doesn’t use my input to train its models. I paste my cleaned text into the AI interface and prompt it to generate an outline. - Outline Refinement
I review the AI-generated outline carefully. I check if the structure makes sense, if the main points are in the right order, and if any key ideas are missing. I make adjustments—reordering sections, adding missing points, or refining headings—until the outline reflects my intended message. This step is crucial because it sets the foundation for the article. - Article Generation
Once I’m satisfied with the outline, I prompt the AI to write a full article based on it. The AI creates a draft that turns my ideas into a coherent narrative, often including connections and explanations I hadn’t thought of. - Article Review & Revision
I then evaluate the AI-generated article. I check for tone, flow, and accuracy—making sure it reflects my voice and intent. If something is missing, too vague, or off-topic, I ask the AI to rewrite it—adding, removing, or rephrasing sections as needed. This may take several iterations, but I keep refining until I’m happy with the result. - Expansion with External Research
Once I have a solid draft, sometimes I prompt the AI to expand it by incorporating relevant external information—such as statistics, expert opinions, or related concepts from credible sources. This helps enrich my original ideas and introduces new perspectives I hadn’t considered. - Final Editing and Publishing
This is where the real work begins. I copy the final article into a word processor or a WordPress draft. Then I go through it line by line, editing it to ensure it sounds like me—adjusting tone, sentence structure, and emphasis to match my voice and style. I don’t just publish it; I treat it as a living document. I often spend days reviewing it for just a few minutes each time with fresh eyes. This usually results in having 4-6 draft articles I’m working on at any given time. Finally I’ll publish it when I’m satisfied. But even after it’s published, I continue to read and refine it, making small improvements over time.
This workflow lets me turn my raw thoughts into polished, publishable content—using AI as a powerful collaborator while keeping full creative control over the end product.
AI Disclosure: The ideas and content of this article are my own but the text was written and refined using artificial intelligence tools. While the core concepts reflect my original thoughts, the phrasing and structure has been optimized by AI.